Hi, I want to ask a question about Cox regression.
I used Cox and Weibull models to regress autocratic regimes' time to failures, my main predictors are two continuous variables (a&b) and their interactions (aXb), and I have other control variables, in total 11 IV.
The regression results are similar, but when I use stcurve to plot the survival probability given different values of a & b, Cox regression cannot plot properly while Weibull can. Here are the exported graphs of them respectively.
Please tell me if you know why this happens, thank you very much.
The codes for Cox regression is
stcox a b aXb ..., vce(cluster c) nohr
No. of subjects = 127 Number of obs = 1,244
No. of failures = 73
Time at risk = 1244
Wald chi2(26) = 151.35
Log pseudolikelihood = -191.36384 Prob > chi2 = 0.0000
I used Cox and Weibull models to regress autocratic regimes' time to failures, my main predictors are two continuous variables (a&b) and their interactions (aXb), and I have other control variables, in total 11 IV.
The regression results are similar, but when I use stcurve to plot the survival probability given different values of a & b, Cox regression cannot plot properly while Weibull can. Here are the exported graphs of them respectively.
Please tell me if you know why this happens, thank you very much.
The codes for Cox regression is
stcox a b aXb ..., vce(cluster c) nohr
No. of subjects = 127 Number of obs = 1,244
No. of failures = 73
Time at risk = 1244
Wald chi2(26) = 151.35
Log pseudolikelihood = -191.36384 Prob > chi2 = 0.0000